Quantum Computing and Quantum RF Cryogenic Microwave Engineering Informational

What is the role of a quantum limited amplifier in achieving high fidelity qubit readout?

A quantum-limited amplifier (QLA) is essential for achieving high-fidelity, single-shot readout of superconducting qubit states. Without a QLA, the first-stage amplifier is typically a HEMT LNA at 4K with noise temperature of 2-5K, requiring integration times of 5-20 μs for >99% readout fidelity. A QLA at the mixing chamber stage adds noise at the fundamental minimum set by quantum mechanics: half a photon at the signal frequency (T_N = hf/2k ≈ 0.12K at 5 GHz), a 20-40× improvement over the HEMT. With 20 dB of QLA gain, the HEMT noise contribution is reduced by a factor of 100 in the noise budget, enabling single-shot readout in 0.1-1 μs with >99.5% fidelity. Two QLA technologies are deployed: Josephson Parametric Amplifiers (JPAs) using a single or few Josephson junctions in a resonant circuit, providing 20 dB gain over 5-20 MHz bandwidth, and Traveling Wave Parametric Amplifiers (TWPAs) using thousands of Josephson junctions in a transmission line, providing 20 dB gain over 2-4 GHz bandwidth. The wide bandwidth of TWPAs enables frequency-multiplexed readout of 10-20 qubits through a single amplifier. Single-shot readout is critical for quantum error correction, where qubit states must be measured rapidly and accurately to determine error syndromes without destroying the quantum information on data qubits.

Category: Quantum Computing and Quantum RF

Updated: April 2026

Product Tie-In: Cryogenic Components, Attenuators, Circulators, Cables

Quantum-Limited Amplification for Qubit Readout

Quantum-limited amplification is one of the enabling technologies that transformed superconducting qubits from laboratory curiosities to practical quantum computing elements. Before QLAs, qubit readout required averaging over thousands of measurements, making real-time feedback and error correction impossible.

Readout Fidelity Analysis

Qubit readout fidelity is the probability of correctly identifying the qubit state (|0⟩ or |1⟩). It depends on the signal-to-noise ratio (SNR) achieved during the measurement time T_meas: Fidelity ≈ 1 - erfc(sqrt(SNR/2))/2. For 99% fidelity: SNR > 6.6 (8.2 dB). For 99.9% fidelity: SNR > 9.5 (9.8 dB). The SNR for dispersive readout: SNR = 4 × eta × n_readout × chi^2 × T_meas / kappa, where eta is the readout chain efficiency (ratio of signal power collected to total emitted), n_readout is the photon number in the readout resonator, chi is the dispersive shift (qubit-state-dependent frequency shift of the resonator), and kappa is the resonator linewidth. System noise is: N_noise = k × T_sys × B × T_meas, where T_sys includes amplifier noise and losses. With HEMT only (T_sys = 10K after accounting for losses): SNR = 8.2 dB requires T_meas ≈ 5 μs. With JPA (T_sys = 0.5K): T_meas ≈ 0.3 μs for the same SNR, a 15× speedup.

JPA vs TWPA Comparison

JPA advantages: simpler fabrication (single junction or SQUID), tunable center frequency (via DC flux bias), lower cost ($2,000-5,000 per device), well-characterized noise performance. JPA disadvantages: narrow bandwidth (5-20 MHz at 20 dB gain), serves only 1-3 qubits per device, requires individual pump calibration. TWPA advantages: broad bandwidth (2-4 GHz, covering the full qubit readout band), serves 10-20+ qubits per device, reduces the total number of amplifier chains needed. TWPA disadvantages: complex fabrication (thousands of junctions with tight parameter uniformity), higher cost ($10,000-50,000), pump leakage management (the strong pump tone can cause issues). For systems with <20 qubits: JPAs are adequate and simpler. For systems with >20 qubits: TWPAs are preferred for scalability. Leading TWPA sources: MIT Lincoln Laboratory (Josephson TWPA, kinetic inductance TWPA), UC Berkeley, and commercial suppliers developing products.

Integration Challenges

(1) Pump tone management: the QLA requires a strong microwave pump (~-70 dBm at the device for JPAs), which can leak through to the qubit and cause dephasing or unwanted transitions. Pump cancellation using a second path with opposite phase suppresses leakage by 20-30 dB. (2) Dynamic range: QLAs have very low P1dB (~-100 to -120 dBm, corresponding to 100-10,000 photons). The readout signal must be in this range, and any spurious signals (qubit drive leakage, pump harmonics) can saturate the amplifier. (3) Impedance environment: QLA gain and noise depend on the impedance seen at the input (determined by the circulator and readout resonator). Impedance variations across the readout band can cause gain ripple and degraded noise performance. Use high-isolation circulators and careful impedance matching at the QLA input.

Common Questions

Frequently Asked Questions

What readout fidelity can be achieved with a quantum-limited amplifier?

State-of-the-art readout fidelity with JPAs/TWPAs: 99.5-99.9% in 200-500 ns measurement time. Google Sycamore achieved 99.5% average readout fidelity across 53 qubits in ~600 ns. IBM Eagle processors achieve >99% in ~1 μs. Without QLA (HEMT only): 95-99% in 1-5 μs. The improvement in both fidelity and speed from QLAs is critical for quantum error correction, which requires >99.5% readout fidelity and fast measurement to keep up with real-time error decoding.

Can I use a cryogenic HEMT without a quantum-limited pre-amp?

Yes, but with reduced performance. HEMT-only readout is simpler to implement and may be adequate for: (1) Early-stage qubit characterization (measuring T1, T2, spectroscopy) where averaging is acceptable. (2) Systems with <10 qubits and relaxed real-time requirements. (3) Applications not requiring quantum error correction. The trade-off is readout speed: HEMT-only requires 5-20 μs per measurement (with averaging for fidelity >99%), while QLA-based readout achieves the same fidelity in 0.2-1 μs, enabling 10× faster quantum circuit execution and real-time feedback.

What happens if the quantum-limited amplifier fails?

If a JPA/TWPA fails (junction critical current degraded, pump port damaged, or trapped flux), the readout chain falls back to HEMT-only performance. The system continues to operate but with degraded readout fidelity and speed. Symptoms: reduced readout contrast (lower SNR), higher measurement error rates, and inability to achieve single-shot readout. Recovery: re-bias or re-pump the JPA (sometimes refluxing by warming above Tc and re-cooling resolves trapped flux issues), replace the device (requires a cryostat warmup cycle, 2-5 days). For production quantum computers, redundant readout chains or on-chip amplifier redundancy are being developed to ensure high availability.

Keep Reading